Vertical and horizontal mlcro-scale gradients of oxygen and sulfide and meiofaunal distributions were examined concurrently in laboratory microcosms to assess the importance of tube/ burrow wall chemistry in generating microhabitat for subsurface meiofauna. These distributions were compared to field distributions at a subtidal site in Corpus Christi Bay, Texas (USA). Results reveal a continuum of microhabitats across oxygen and sulfide gradients in marine sands. The majority of taxa llved at [ 0 2 ] below 50 BM. These included microoxyphilic oxybiota living at [O,] >O and thiobiota at [O,] = 0. All classic thiobiota lived below the zero-oxygen line where sedirnents usually contained measurable sulfide. Most, but not all, subsurface taxa were attracted to burrows/tubes; however, only 1 taxon, Praeaphanostoma, lived in the microoxyphilic habitat of the burrow/tube wall's oxic halo. The remainder lived outside this zone where [ 0 2 ] = 0. No dominant species had substantially overlapping distributions. Each abundant species occupied its own unique microhabitat. Most microhabitat boundaries could be explained by changes in 0, and sulfide concentration; however, a few taxa were differentially distributed in microhabitats of similar chemistry, suggesting that other factors were important. Changing sediment chemistry may remove or create optimal microhabitat seasonally so that taxa must be able to exploit a range of suboptimal habitats to survive throughout the year. Most microhabitats, in shortest dimension, were no more than 2 to 3 times the animal's size, suggesting that traverses of no more than a few body lengths would be necessary to move from optimal to suboptimal habitat. Overall, the complex distributional patterns of meiofaunal species suggest that Fenchel & Riedl's (1970) simple 2-layered model of marine sand bottoms, comprising a surface oxic layer and a sulflde system, is inadequate. Even taxa living where [02] = 0 were distnbuted differentially along the sulfide gradient. Whether thiobiota and oxybiota are merely 2 ends of a continuum or really ecologically relevant subsets of subsurface meiofauna must b e reconsidered.